Selenoprotein T is a novel OST subunit that regulates UPR signaling and hormone secretion

Abstract

Selenoprotein T (SelT) is a recently characterized thioredoxin-like protein whose expression is very high during development, but is confined to endocrine tissues in adulthood where its function is unknown. We report here that SelT is required for adaptation to the stressful conditions of high hormone level production in endocrine cells. Using immunofluorescence and TEM immunogold approaches, we find that SelT is expressed at the endoplasmic reticulum membrane in all hormone-producing pituitary cell types. SelT knockdown in corticotrope cells promotes unfolded protein response (UPR) and ER stress and lowers endoplasmic reticulum-associated protein degradation (ERAD) and hormone production. Using a screen in yeast for SelT-membrane protein interactions, we sort keratinocyte-associated protein 2 (KCP2), a subunit of the protein complex oligosaccharyltransferase (OST). In fact, SelT interacts not only with KCP2 but also with other subunits of the A-type OST complex which are depleted after SelT knockdown leading to POMC N-glycosylation defects. This study identifies SelT as a novel subunit of the A-type OST complex, indispensable for its integrity and for ER homeostasis, and exerting a pivotal adaptive function that allows endocrine cells to properly achieve the maturation and secretion of hormones.

Keywords: ER stress; KCP2; N‐glycosylation; oligosaccharyl transferase; pituitary.

Figure 1. Alteration of adrenocoticotropic hormone secretion in SelT‐deficient corticotrope cells

1. Double immunostaining of pituitary sections for SelT (green) and hormones (red). Nuclei were labeled with DAPI (blue). Each hormone positive cell type expresses SelT (white arrows), but not all cells (yellow arrows) contain SelT. Scale bar, 12 μm.

2. TEM images of adenohypophysis showing SelT ultrastructural distribution. Sections were treated with SelT antibody and visualized with silver‐enhanced, gold‐labeled secondary antibodies. (a, b) Heavily labeled cells revealed by silver grains (arrowheads) are adjacent to weakly labeled ones (asterisks) characterized by distinct secretory content (arrows). (c, d) Immunopositive cells display silver grains on the ER membrane (arrowheads) and at the plasma membrane (double arrowhead in d). (e‐g) Labeling (arrowheads) is restricted to cytoplasmic domains excluding secretory granules (SG) and mitochondria (M). Scale bar, 0.5 μm.

3. Schematic representation of POMC, which generates ACTH after cleavage by PC1 in AtT20 cells.

4. Western blot analysis of SelT expression in mouse pituitary and pituitary corticotrope AtT20 cells.

5. Immunohistochemical analysis of Myc‐SelT (red) intracellular distribution in AtT20 cells; the ER compartment was labeled with transiently expressed ER‐GFP (green). DAPI was used to stain the nuclei. Scale bar, 12 μm.

6. A Myc‐SelT variant carrying a deletion of the hydrophobic domains (SelTΔTD) showed relocation from the ER (left) to the cytosol (right). Scale bar, 12 μm.

7. qRT–PCR analysis of SelT expression in the presence of CRF (100 nM, 24 h) in scr shRNA‐transduced AtT20 cells. Asterisks denote statistical significance. Each value is the mean ± SEM (n = 7, *P < 0.05; Mann–Whitney test).

8. Analysis of ACTH concentration in cell culture supernatants by ACTH ELISA assay. Scr and SelT shRNA‐transduced AtT20 cells were treated with 100 nM of CRF for 24 h. Data are expressed as percentage of control (untreated cells; set to 100). Each value is the mean ± SEM (n = 3, *P < 0.05; Mann–Whitney test).

9. Analysis of scr and SelT shRNA‐transduced AtT20 cells content by ACTH ELISA assay. Data are expressed as percentage of control (untreated AtT20 cells; set to 100). Each value is the mean ± SEM (n = 3; *P < 0.05; **P < 0.01, two‐way ANOVA).

Source data are available online for this figure.

Figure 2. SelT is required for ER homeostasis

1. Confocal imaging of ER in scr and SelT shRNA‐transduced AtT20 cells transfected with ER‐DsRed vector (red) and grown at 37 or 39°C for 16 h. Nuclei were labeled with DAPI (blue). The number of red pixels per cell, representing the ER area, was measured using the ImageJ software. Data are expressed as percentage of control (scr transduced cells grown at 37°C; set to 100). Scale bar, 25 μm. Each value is the mean ± SEM (n = 8; *P < 0.05; **P < 0.01; ***P < 0.0001; Mann–Whitney test).

2. Western blot analysis of ER stress markers, CHOP and BIP, in scr and SelT shRNA‐transduced AtT20 cells. GAPDH was used as an internal loading control. Relative protein levels were measured with the Image Lab software (Bio‐Rad) and expressed as percentage of control (scr set to 100). Each value is the mean ± SEM (n ≥ 3; **P < 0.01; ***P < 0.001; Mann–Whitney test).

3. qRT–PCR analysis of the expression of ER stress‐responsive genes in scr and SelT shRNA‐transduced AtT20 cells. Data are expressed as percentage of control (scr set to 100). Each value is the mean ± SEM (n ≥ 3; *P < 0.05; **P < 0.01; Mann–Whitney test).

4. qRT–PCR analysis of the expression of ER stress‐responsive genes in scr and CRF (100 nM, 24 h)‐treated scr cells. Each value is the mean ± SEM (n ≥ 2, *P < 0.05; **P < 0.01; Mann–Whitney test).

5. Western blot analysis of SelT and BIP expression in AtT20 cells after Bip knock down in basal conditions and after CRF (100 nM) treatment for 24 h.

6. qRT–PCR analysis of SelT expression in the presence of two ER stressors, Tuni (1 μg/ml) and DTT (100 μM) for 6 h in wild‐type (scr) AtT20 cells (Ctrl set to 100). Each value is the mean ± SEM (n = 3–5; *P < 0.01; one‐way‐ANOVA).

7. Comparison of the viability of scr and SelT shRNA‐transduced AtT20 cells treated with Tuni (1 μg/ml) or DTT (100 μM) for 24 h. Data are expressed as percentage of control (scr set to 100). Each value is the mean ± SEM (n = 5; *P < 0.01; **P < 0.01; one‐way‐ANOVA).

8. Schematic representation of shRNA‐resistant forms of SelT, native, or carrying a series of substitutions in the CSVU motif, designed to rescue the SelT shRNA effect.

9. Detection of recombinant SelT mutants in SelT shRNA‐expressing AtT20 cells by Western blot analysis.

10. Viability analysis of scr and SelT shRNA‐transduced AtT20 cells rescued with a series of mutants described in (H), in ER stress conditions using 1 μg/ml Tuni (top panel) or 100 μM DTT (bottom panel) for 24 h. Each value is the mean ± SEM (n ≥ 5; *P < 0.05; **P < 0.01; ***P < 0.001; Mann–Whitney test).

Source data are available online for this figure.

Figure 3. SelT deficiency slows down ER‐associated protein degradation

1. Scr and SelT shRNA‐transduced AtT20 cells were transiently transfected with an expression plasmid encoding the GFP‐tagged Hong Kong variant of mutant α1‐antitrypsin (NHK‐GFP). AtT20 cells were pulse‐chased for the indicated time and analyzed by Western blot analysis using an anti‐GFP antibody. α‐tubulin was used as an internal loading control.

2. Quantification of GFP‐tagged NHK degradation rate in (A). Relative NHK‐GFP levels were calculated by dividing the normalized GFP signals in cells after 1.5, 3, and 4.5 h of cycloheximide (CHX) treatment with that of untreated cells (untreated cells set to 100). Each value is the mean ± SEM (n = 4; **P < 0.01; ***P < 0.001; two‐way ANOVA test).

Source data are available online for this figure.

Figure 4. SelT interacts with KCP2 and stabilizes A‐type OST subunits

1. ER membrane topology of KCP2 (generated with Protter http://wlab.ethz.ch/protter/start/).

2. Schematic representation of the four KCP2‐related prey fragments found to interact with SelT in the yeast two‐hybrid screen. In green, lumenal domains; in orange, cytoplasmic domains.

3. Co‐immunoprecipitation analysis of SelT and KCP2 interactions in AtT20 cells. Proteins interacting with endogenous SelT and KCP2 were immunoprecipitated with SelT and KCP2 antibodies, respectively, and analyzed by Western blot. IgG, irrelevant antibody.

4. Co‐immunoprecipitation analysis of SelT and STT3A, STT3B, and OST48 subunit interactions in AtT20 cells. Proteins interacting with endogenous SelT, STT3A, STT3B, OST48, and KCP2 were immunoprecipitated with SelT, STT3A, STT3B, OST48, and KCP2 antibodies, respectively, and analyzed by Western blot. IgG, irrelevant antibody.

5. Dual immunofluorescence staining showing SelT (green) and KCP2 (red) colocalization in AtT20 cells. Scale bar, 10 μm.

6. Western blot detection of SelT and KCP2 in AtT20 cell microsomal preparation. SelT and KCP2 were detected in whole AtT20 cell lysate (Ctrl) and in the microsomal (Mic) fraction.

7. Western blot analysis of SelT expression in control AtT20 cells in the presence or absence of Tuni (1 μg/ml). GAPDH was used as an internal loading control.

8. Western blot analysis of SelT expression in scr and SelT shRNA‐transduced AtT20 cells. α‐tubulin antibody was used as a loading control. Data are expressed as percentage of control (scr; set to 100). Each value is the mean ± SEM (n = 5; **P < 0.01; t‐test).

9. KCP2, STT3A, OST48, STT3B, and Sec61β protein levels were analyzed by Western blot in scr and SelT shRNA‐transduced AtT20 cells. GAPDH was used as an internal loading control. Relative protein levels were measured with the Image Lab software (Bio‐Rad) and expressed as percentage of control (scr; set to 100) (right panel). Each value is the mean ± SEM (n ≥ 3; *P < 0.05; **P < 0.01; Mann–Whitney test).

10. qRT–PCR analysis of KCP2, STT3A, and OST48 mRNA levels in scr and SelT shRNA‐transduced AtT20 cells. Data are expressed as percentage of control (scr; set to 100). Values are expressed as mean ± SEM (n ≥ 3; Mann–Whitney test).

11. KCP2, SelT, STT3A, and OST48 protein levels were analyzed by Western blot in scr and KCP2 siRNA‐tranduced AtT20 cells. α‐tubulin was used as an internal loading control. Relative protein levels were measured with the Image Lab software (Bio‐Rad) and expressed as percentage of control (scr; set to 100) (right panel). Each value is the mean ± SEM (n ≥ 3; *P < 0.05; Mann–Whitney test).

Source data are available online for this figure.

Figure 5. SelT depletion perturbs N‐glycosylation

1. Schematic representation of the Glyc‐ER‐GFP construct. Red star: Mutation N147T inserted; ER signal: calreticulin ER signal sequence; KDEL: ER retention sequence (modified from Losfeld et al, 2012).

2. Scr and SelT shRNA‐tranduced AtT20 cells were transiently transfected with ER‐GFP and Glyc‐ER‐GFP (green). Glyc‐ER‐GFP transfected cells were immunolabeled by anti‐GFP (red). For Glyc‐ER‐GFP quantification, the number of green and red pixels was measured using the ImageJ software. Number of green pixels was normalized to the number of red pixels in each condition. Nuclei were labeled with DAPI (blue). Scale bar, 50 μm. Data are expressed as percentage of control (scr set to 100). Each value is the mean ± SEM (n = 4; *P < 0.05; Mann–Whitney test).

3. Western blot analysis of expression of glycosylated or non‐glycosylated GFP in scr and SelT shRNA‐tranduced AtT20 cells transiently transfected by Glyc‐ER‐GFP and treated or not with 1 μg/ml of Tuni.

4. Diagram of POMC precursor showing the disulfide bridges and the N‐glycosylation sequons.

5. Western blot analysis of glycoforms of exogenous Flag‐POMC in AtT20 cells, after SelT gene disruption using CRISPR Cas9 technology, or in the presence of Tuni. Glycoforms in the cells invalidated for the SelT gene by CRISPR Cas9 technology were expressed relative to the normalized signal in non‐invalidated cells. Quantified values below gel lanes are for the displayed image, that is, representative of two or more experiments.

Source data are available online for this figure.